1. Field of the Invention
The invention relates to a method for selecting a locality name in a motor vehicle navigation system by voice input. Furthermore, the invention relates to a navigation system comprising a voice input system for converting spoken words into features or, respectively, a sequence of features, a non-volatile storage element as a first data .source which contains a list with locality names, a second data source which contains some of the data of the first data source, a control unit for selecting a locality name on the basis of the sequence of features supplied by the voice input system.
2. Background of the Invention
Navigation systems for motor vehicles have been known for a relatively long time and are increasingly offered in new vehicles or for retrofitting. Such navigation systems contain a storage medium containing the road map data of one or more countries. Before starting a voyage, the user can have a route calculated to his destination and then allow the navigation system to guide him to the selected destination. In the navigation systems currently offered, the destination is entered manually, as a rule. In addition, however, navigation systems are also already known in which a destination is selected by voice input. The problem with this arrangement is, however, that several 10 000 town names and several 1,000 street names must be considered as locality names for the destination input in a navigation system, depending on the size of the geographic area stored in the storage medium. Known voice recognition devices, in contrast, are designed for recognizing much fewer words, for example about 1,000. In addition, searching in such a large database takes a correspondingly long time.
From DE 197 42 054 A1, an input system for town and/or street names is known which contains a data source which contains, apart from a first town and/or street list containing alphabetically sorted town and/or street names, at least one second town and/or street list with town and/or street names sorted in accordance with a frequency criterion. The frequency criterion selected is, in particular, the number of inhabitants of a town or the number of inhabitants living in a street. The search for a locality name which has been input is initially performed in the second list and is only performed in the first complete list if the town which has been input has not been found in the second list. As a result, large cities, for example, given as destinations are very rapidly found. The locality names can be input via a voice input system which, however, is not described in greater detail.
From DE 197 09 518 C1, a method and a device for inputting a destination address by voice into a navigation system in real-time mode is known. The voice recognition system exhibits both a speaker-independent voice recognizer and a speaker-dependent voice recognizer, the speaker-independent voice recognizer operating on the basis of phonemes. To ensure that voice can be recognized in real time, the address data are allocated to various lexicons. Such a lexicon then always represents a subset of all stored address data and has a maximum of 1,500 entries. A basic lexicon is provided which has approximately 1,000 entries of the towns having more than 10,000 inhabitants. In addition, lexicons for individual regions or surrounding areas are provided. However, this method is very costly due to the multiplicity of lexicons and the requirement for performing two passes in the voice recognition system. In addition, the data must already be correspondingly edited and allocated to various lexicons.
The purpose and advantages of the present invention will be set forth in and apparent from the description that follows, as well as will be learned by practice of the invention. Additional advantages of the invention will be realized and attained by the methods and systems particularly pointed out in the written description and claims hereof, as well as from the appended drawings.
It is an object of the invention to specify a method for selecting a locality name in a motor vehicle navigation system with voice input which allows a fast selection of localities by voice input of the locality names already previously selected by a user. It is a further object of the invention to develop a generic navigation system in such a manner that it is suitable for carrying out this method.
In brief, an embodiment of the invention is a method for selecting a locality name in a motor vehicle navigation system in which said system contains a first and a second list with locality names, wherein said second list comprises a subset of the locality names contained in the first list together with the phonemes of said locality names, comprising:
vocally inputting a desired locality name via a voice input system;
determining features of the vocally input desired locality name;
comparing the features of the vocally input desired locality name with features of the stored phonemes of the locality names in the second list; and
selecting a locality name if the features of the vocally input desired locality name sufficiently correspond to the features of the phonemes of a locality name in the second list; or
if the features of the vocally input desired locality name do not sufficiently correspond to the features of the phonemes of a locality name in the second list, requesting a further input of the desired locality name in a second form of input differing from the voice input of the desired locality name;
selecting a locality name from the first list by means of the second form of input; and
storing the selected locality name and the phonemes of the vocally input desired locality name in the second list.
The comparison between the features of the spoken locality name and the features of the stored phonemes is performed with the aid of known methods for pattern comparison.
The method according to the invention uses a speaker-independent voice input system so that no special training for voice recognition is required. All locality names which are available in the associated storage element are stored in a first list. In this connection, a locality name is understood to be both town names and street names. If necessary, the first list can be divided into two part-lists, namely one for the town names and one for street names. In the second list, only those locality names are entered which have already been entered into the system by voice input by a user. In this second list, both the locality name in written form and a stylized form of the spoken locality name in the form of phonemes are stored for this purpose. If then a repeated voice input of a locality name takes place via the voice input unit, features or, respectively, a sequence of features of the spoken locality name are first determined in a manner known per se. Then these features of the spoken locality name are compared with the features of the phonemes of locality names stored in the second list. If sufficient correspondence to a stored locality name is found, this locality name is selected. Due to a relatively small number of locality names in the second list, this selection is carried out in real time.
If it has not been possible to find in the second list the locality name which has been input, the user is requested to input the locality name in a second form of input. This second form of input can be, for example, the manual input of the locality name or a voice input in the form of numbers or letters. In the latter case, the user spells the locality name, for example, or he inputs the telephone dialing code or zip code of the town to be found. The locality name is selected from the first complete list by using this second input form. The selected locality name is then stored, together with the phonemes of the locality name previously spoken, in the second list so that it can be found in the case of a repeated voice input of this locality name. The locality names of the first list can be associated with data records which contain, for example, more detailed information on this town. In a special embodiment, it is provided that the data records which are associated with a locality name of the first list, after a locality name has been selected from the first list by means of the second form of output are stored, together with the selected locality name, in the second list. Thus, the information of these data records is also immediately available in the case of a repeated selection of this town by voice input.
The locality names can be sorted in different ways, especially in the second list. On the one hand, the locality names can be stored alphabetically. On the other hand, however, the navigation system can be adapted better to the habits of the respective user if the frequency of input of a locality name is additionally recorded. In this case, the locality names of the second list can be sorted in accordance with the frequency in which they have been input. The advantage of this is that the locality name which is selected most frequently by the respective user is at the top in the second list. It will thus also be found most rapidly in the case of a repeated input. This results in a particularly short response time of the system for the towns frequently input and thus also driven to by the user.
In a further embodiment of the invention, it is provided that the locality names of the second list are stored in the order in which they have been input. As a result, the trip destinations last driven to are in each case at the top in the second list. Such a method is of advantage if individual destinations are driven to more frequently only in particular time intervals. In addition, combinations of the sorting criteria described above are also possible. Thus, for example, a weighted storage of the locality names in accordance with the frequency in which they have been input and the order in which they have been input is conceivable.
In brief, yet another embodiment of the invention is a navigation system for a motor vehicle comprising a voice input system for converting spoken words into a plurality of features; a first data source comprising a non-volatile storage element for storing data comprising a list with locality names; a second data source for storing some of the data of the first data source; and a control unit for selecting a locality name on the basis of the features supplied by the voice input system. Further embodiments are provided wherein the second data source comprises a rewritable non-volatile storage element for storing both locality names and phonemes of the locality names associated with the locality names, wherein said phonemes are generated by the voice input system on the basis of the spoken words.
The navigation system preferably has a manual input device for inputting locality names in a second form of input. This is of advantage, in particular, if the locality name cannot be identified by voice input.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and are intended to provide further explanation of the invention claimed.
The accompanying drawings, which are incorporated in and constitute part of this specification, are included to illustrate and provide a further understanding of the method and system of the invention. Together with the description, the drawings serve to explain the principles of the invention.